The structure of GDP-4-keto-6-deoxy-d-mannose-3 ... · pyridoxamine 59-phosphate (PMP) by reacting with gluta-mate as indicated in Scheme 2. In the first step, glutamate forms a Schiff

The structure of GDP-4-keto-6-deoxy-D-mannose-3-dehydratase: A unique coenzyme B6-dependent enzyme

PAUL D. COOK, JAMES B. THODEN, AND HAZEL M. HOLDENDepartment of Biochemistry, University of Wisconsin—Madison, Madison, Wisconsin 53706, USA

(RECEIVED May 3, 2006; FINAL REVISION June 12, 2006; ACCEPTED June 18, 2006)

Abstract

L-Colitose is a 3,6-dideoxysugar found in the O-antigens of some Gram-negative bacteria such asEscherichia coli and in marine bacteria such as Pseudoalteromonas tetraodonis. The focus of thisinvestigation, GDP-4-keto-6-deoxy-D-mannose-3-dehydratase, catalyzes the third step in colitoseproduction, which is the removal of the hydroxyl group at C39 of GDP-4-keto-6-deoxymannose. It isan especially intriguing PLP-dependent enzyme in that it acts as both a transaminase and a dehydratase.Here we present the first X-ray structure of this enzyme isolated from E. coli Strain 5a, type O55:H7.The two subunits of the protein form a tight dimer with a buried surface area of ;5000 A2. This isa characteristic feature of the aspartate aminotransferase superfamily. Although the PLP-binding pocketis formed primarily by one subunit, there is a loop, delineated by Phe 240 to Glu 253 in the secondsubunit, that completes the active site architecture. The hydrated form of PLP was observed in one of theenzyme/cofactor complexes described here. Amino acid residues involved in anchoring the cofactor tothe protein include Gly 56, Ser 57, Asp 159, Glu 162, and Ser 183 from one subunit and Asn 248 fromthe second monomer. In the second enzyme/cofactor complex reported, a glutamate ketimineintermediate was found trapped in the active site. Taken together, these two structures, along withpreviously reported biochemical data, support the role of His 188 as the active site base required forcatalysis.

Keywords: PLP-dependent enzyme; colitose; molecular structure; 3,6-dideoxysugars; O-antigens

The biochemical importance of carbohydrates cannot beoverstated, for they are essential elements in nearly everyphysiological process and represent the most abundantbiomolecules in living systems. Di- and trideoxysugarsare an especially important and intriguing class ofcarbohydrates that are synthesized by plants, fungi, andbacteria (Kennedy and White 1983; Rupprath et al. 2005).Many are formed from simple monosaccharides such asglucose 6-phosphate or fructose 6-phosphate via a myriadof enzymatic reactions including dehydrations, reduc-tions, and methylations. Most of the deoxysugars of

biological relevance are constructed around the 3,6-dideoxyhexoses or the 2,3(4),6-trideoxyhexoses, whichare further modified by epimerization, amination, and C-,N-, and O-methylation reactions. These modificationsyield an astonishing array of biological molecules withdiffering three-dimensional shapes and hydrophilicities.

Several of these unusual sugars have been isolated fromthe O-antigens of various Gram-negative bacteria, wherethey contribute to the enormous structural variations ob-served in bacterial cell walls (Lerouge and Vanderleyden2002). Others have been found in polyketide antibioticssuch as erythromycin, where they have been shown toprovide or enhance the pharmacological activities of therespective drugs (Weymouth-Wilson 1997).

One of these unusual sugars is L-colitose (Scheme 1),a 3,6-dideoxysugar isolated from the O-antigens of Gram-negative bacteria such as Escherichia coli (Edstrom andHeath 1965), Yersinia pseudotuberculosis (Kotandrova

ps0623283 Cook et al. ARTICLE RA

Reprint requests to: Hazel M. Holden, 433 Babcock Drive, Depart-ment of Biochemistry, University of Wisconsin––Madison Madison,WI 53706, USA; e-mail: [email protected]; fax: (608)262-1319.

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.062328306.

Protein Science (2006), 15:2093–2106. Published by Cold Spring Harbor Laboratory Press. Copyright � 2006 The Protein Society 2093

JOBNAME: PROSCI 15#9 2006 PAGE: 1 OUTPUT: Monday August 7 15:31:19 2006

csh/PROSCI/118160/ps0623283

et al. 1989), Salmonella enterica (Xiang et al. 1993),Vibrio cholerae (Hisatsune et al. 1993), and in marine bac-teria such as Pseudoalteromonas tetraodonis (Muldoonet al. 2001) and Pseudoalteromonas carrageenovora(Silipo et al. 2005). Reports describing the isolation ofcolitose date as far back as 1958 (Luderitz et al. 1958).

The biosynthesis of the O-antigen in bacteria requiresnucleotide-linked sugars, which are transferred to thegrowing oligosaccharide chain. Four enzymes are in-volved in the production of GDP-L-colitose, as outlinedin Scheme 1 (Bastin and Reeves 1995; Beyer et al. 2003).The focus of this investigation is GDP-4-keto-6-deoxy-D-mannose-3-dehydratase, which catalyzes the third step inthe pathway, namely the removal of the hydroxyl groupat C39 of the hexose. The enzyme is encoded by the colDgene, and hereafter is referred to as ColD. It is anespecially intriguing PLP-dependent enzyme in that itlacks the conserved lysine that typically anchors thecofactor to the protein. Recent biochemical studies havedemonstrated that the enzyme from Y. pseudotuberculosisis bifunctional, with both deoxygenation and transamina-tion activities (Beyer et al. 2003; Alam et al. 2004),and as such, it represents a new paradigm for unusualdeoxysugar formation.

PLP (pyridoxal 59-phosphate) is a remarkable andversatile coenzyme that is ubiquitous in nature. Enzymesdependent upon PLP catalyze a wide range of chemicalreactions including decarboxylations, racemizations, trans-aminations, and b-eliminations, among others (Jansonius1998; Toney 2005). Most PLP-dependent enzymes contain

a conserved lysine residue, which forms a Schiff base withthe coenzyme in the resting state. This species is referred toas the internal aldimine. In ColD, an internal aldimine isnot possible due to the replacement of a histidine for theconserved lysine. Rather, in ColD, PLP is first converted topyridoxamine 59-phosphate (PMP) by reacting with gluta-mate as indicated in Scheme 2. In the first step, glutamateforms a Schiff base with the noncovalently bound PLP,a species referred to as the external aldimine. Subsequently,an enzymatic base deprotonates the aldimine intermediate,leading to the formation of a carbanion that is stabilized byvarious resonance forms. Protonation of the carbanion atC49 of the PLP by an active site acid leads to a ketimineintermediate which, upon hydrolysis, results in PMP anda-ketoglutarate.

Here we describe the three-dimensional structures ofColD from Escherichia coli, Strain 5a, type O55:H7, witha bound hydrated form of PLP and with a trappedglutamate ketimine intermediate. The employment ofPMP in a dehydration reaction is quite rare, and thusthese structures provide a wealth of new molecular insightinto this novel coenzyme B6-dependent enzyme.

Results

The first structure of ColD was determined to 1.8 Aresolution. For this structural analysis, crystals weregrown in the presence of PLP and 2 mM a-ketoglutarate.As can be seen in Figure 1A, the two subunits of ColDform a tight dimer with an extensive total buried surfacearea of ;5000 A2, based on a probe radius of 1.4 A(Zhang and Matthews 1995). This type of quaternarystructure is characteristic of the aspartate aminotransfer-ase superfamily (Jansonius 1998). The dimer has overalldimensions of ;80 A 3 70 A 3 70 A, with the activesites separated by ;20 A. Electron density correspondingto the entire polypeptide chain backbone is continuousfrom the N to C termini for both monomers in theasymmetric unit. The conformations of the two subunitsforming the dimer are virtually identical, and correspondwith a root-mean-square deviation (RMSD) of 0.14 A for367 structurally equivalent a-carbons.

A ribbon representation of an individual subunit (a totalof 388 amino acids) is displayed in Figure 1B. It hasa somewhat concave-shaped appearance with moleculardimensions of ;60 A 3 65 A 3 70 A. The polypeptidechain initiates with two a-helices defined by Asp 13–Ser25 and Glu 32–Phe 45. These helices lead up to the firststrand of b-sheet labeled ‘‘A’’ in Figure 1B. Overall, thefold of the subunit is dominated by a mixed eight-stranded b-sheet flanked by eight a-helices. Theb-strands are labeled ‘‘A,’’ ‘‘G,’’ ‘‘F,’’ ‘‘E,’’ ‘‘D,’’ ‘‘B,’’‘‘C,’’ and ‘‘I’’ in Figure 1B, and correspond to Tyr 49–Val53, Gly 196–Thr 200, Met 179–Ser 184, Ile 155–Asp 159,

Scheme 1.

Cook et al.

2094 Protein Science, vol. 15



Scheme 1 live 4/C

Lys 128–Asn 134, Glu 80–Pro 84, Arg 101–Val 105, andThr 351–His 353, respectively. Note that b-strands A, G,and F run antiparallel. In addition to this large b-sheet,there are two strands of antiparallel sheet labeled ‘‘H’’and ‘‘J’’ that correspond to residues Gly 303–Ile 308, andLeu 367–Val 369, respectively. A rather large helix,formed by Glu 253–Lys 283, packs against both the firsthelix at the N terminus and the last helix at the Cterminus. The third helix of the subunit, which connectsb-strands A and B, is positioned such that the positive endof its helix dipole moment projects toward the phosphateof the PLP cofactor. Those residues involved in anchoringthe PLP ligand to the protein are provided primarily byone subunit. However, as shown in Figure 1C, there isa loop, delineated by Phe 240–Glu 253, which forms onewall of the active site and is provided by the secondsubunit of the dimer.

Unlike the typical PLP-dependent enzymes, ColD doesnot contain the lysine residue necessary to form a Schiffbase between its side chain Nz and C49 of the cofactor

(the internal aldimine). Aldehydes, like PLP, react withwater to form hydrates, as indicated in Figure 2A.Whether the carbonyl or hydrate moiety is favored atequilibrium depends upon the chemical nature of thealdehyde (or ketone). In the case of free PLP in solution,the aldehyde form is favored. There is spectral evidence,however, that when PLP is bound to a protein, thehydrate may be favored (Harris et al. 1976; Nishino etal. 1997; Tramonti et al. 2002). Shown in Figure 2B is theelectron density corresponding to the PLP ligand insubunit 2. Originally the ligand was modeled into theelectron density as an aldehyde with C49 having sp2

hybridization. As the model refinement continued, how-ever, it became clear from electron density maps calculatedwith (Fo � Fc) coefficients that, in fact, the hydrated formof PLP, with C49 having sp3 hybridization, was trapped inthe active site.

Those residues located within ;3.5 A of the hydratedPLP in subunit 2 are shown in Figure 2C. Most of theprotein:ligand interactions occur near the phosphoryl

Scheme 2.

Structure of ColD

www.proteinscience.org 2095



Scheme 2 live 4/C

Figure 1. The structure of ColD. The crystals of ColD employed in this investigation contained a dimer in the asymmetric unit as shown in

A. The subunits are related by a twofold rotational axis, which lies perpendicular to the plane of the page. The PLP cofactors, separated by

;21 A, are displayed in space-filling representations. For all figures, the coordinates corresponding to subunits 1 and 2 of the dimer are color

coded in blue and green, respectively. A separate ribbon representation of subunit 2 is given in B. The 10 b-strands characterizing the subunit

are labeled A–J. Part of the active site region for subunit 2 is formed by a loop contributed by subunit 1 (Phe 240–Glu 253) as highlighted in C.

Cook et al.




Fig. 1 live 4/C

group. Indeed, the phosphoryl moiety is anchored firmlyin place by eight hydrogen bonding interactions. Specif-ically, the backbone amide groups of Gly 56 and Ser 57

participate in hydrogen bonding interactions with two ofthe phosphoryl oxygens of PLP. In addition, the sidechain atoms, Og of Ser 57 (subunit 2) and Nd2 of Asn 248

Figure 2. The complex of ColD with the hydrated form of PLP. Aldehydes react with water to form hydrates. This type of reaction for

PLP is shown in A, along with the numbering used to define the various atoms of the cofactor. Electron density corresponding to the

hydrated form of PLP is displayed in B. The map, contoured at 2s, was calculated with coefficients of the form (Fo � Fc), where Fo

was the native structure factor amplitude and Fc was the calculated structure factor amplitude from the model lacking coordinates for

the PLP. Those residues situated within ;3.5 A of the hydrated PLP in subunit 2 are shown in C. The dashed lines indicate possible

hydrogen bonds between atoms positioned within 3.2 A of each other. Asn 248 is contributed by subunit 1.

Structure of ColD




Fig. 2 live 4/C

(subunit 1), lie within 2.4 A and 2.9 A, respectively, ofone of the phosphoryl oxygens. Ser 183 provides anotherhydrogen bonding interaction between its side chainOg and a PLP phosphoryl oxygen. Three ordered watermolecules complete the hydrogen bonding pattern ob-served around the phosphoryl group. Only two sidechains, Asp 159 and Glu 162, form hydrogen bonds withthe pyridoxal ring of PLP. Asp 159 is positioned at theend of b-strand E, whereas Glu 162 resides in a Type Iturn. For the side chain oxygen of Glu 162 to lie within;2.8 A of O39 of the pyridoxal ring, one of these oxygensmust be protonated. The indole ring of Trp 88 provides anoff-centered parallel stacking interaction with the co-factor. Note that His 188, which is a lysine residue inmost PLP-dependent enzymes, is located at ;4 A fromthe cofactor C49.

For the second structure reported here, ColD wascrystallized in the presence of 2 mM glutamate and thestructure solved to 1.9 A resolution. Again, the twosubunits in the asymmetric unit are virtually identical,and correspond with an RMSD of 0.18 A for 372structurally equivalent a-carbons.

A close-up view of the electron density observed in theactive site of subunit 2 is displayed in Figure 3A. Initiallythe external aldimine form of PLP was modeled intothe electron density whereby the a-amino nitrogen ofglutamate and C49 of the pyridoxal ring form a Schiffbase. For this intermediate, the hybridization about thea-carbon of glutamate is sp3 while the hybridizationabout C49 of the cofactor is sp2 (Scheme 2). Accordingly,the bond angles about the a-carbon and C49 should be;109° and ;120°, respectively. During the course of thestructure refinement, however, the external aldiminemodel never matched the electron density well. If,however, the hybridization about the a-carbon and C49were restrained to sp2 and sp3, respectively, the modelfitted strikingly well into the electron density (Fig. 3A).This type of geometry at the a-carbon and C49 would beexpected for the glutamate ketimine intermediate(Scheme 2). While caution must be applied when inter-preting electron density maps calculated at 1.9 A resolu-tion, the observation of such an intermediate is notunprecedented. Indeed, it has been previously reportedin the elegant structural analyses of aspartate aminotrans-ferase, which were conducted at 2.3–2.4 A resolution(Malashkevich et al. 1993). Aspartate aminotransferasecontains the typical lysine residue observed in most PLP-dependent enzymes.

Those residues lying within 3.5 A of the glutamateketimine intermediate in subunits 1 and 2 are displayed inFigure 3B. Note that the a-carbon chains for the ColDdimer, whether complexed with either the hydrated orketimine form of PLP, are virtually identical (RMSD of0.09 A). The protein accommodates the more bulky

ketimine intermediate by the exclusion of water mole-cules from the active site rather than by major side chainadjustments. Both His 215 and Arg 250 from the firstsubunit form electrostatic interactions with the side chaincarboxylate group of the ketimine intermediate. Asobserved for the protein complexed with the PLP hydrate,one of the carboxylate oxygens of Glu 162 lies quiteclosely to O39 of the pyridoxal ring (;2.6 A), which isindicative of a proton being shared between them.

Discussion

The structure of ColD was solved via molecular re-placement with the X-ray coordinates for ArnB fromSalmonella typhimurium (Noland et al. 2002). These twoenzymes demonstrate a 23% amino acid sequence iden-tity. ArnB plays a key role in lipid A modification bycatalyzing the amination at C-40 of UDP-L-threo-penta-pyranosyl-40-ulose. Both ColD and ArnB are dimeric.Their structurally equivalent a-carbons correspond withan RMSD of 2.5 A. A superposition of one subunit ofColD onto that of ArnB is displayed in Figure 4A. As canbe seen, the overall backbone traces for the two enzymesare very similar. They begin to align at Leu 6 in ColD andPhe 13 in ArnB. The two proteins differ primarily in threeareas labeled R1, R2, and R3 in Figure 4A. The firstregion of difference occurs in the loop defined by Phe 68to Arg 74, which serves to connect the third a-helix of theColD subunit and b-strand B. The second area occursnear Lys 224 to Val 230 where, in ColD, these residuesadopt a b-hairpin motif. This region is apparently moreflexible in ArnB where there is a break in the model fromAsp 220 to Gln 232. Finally, the loops leading out ofb-strands H in ColD (Lys 309 to Ile 315) and ArnB adoptsignificantly different orientations (Fig. 4A). In ArnB,Phe 330 is in the cis peptide conformation. The structur-ally equivalent residue in ColD is Thr 334, which adoptsstrained dihedral angles (f ¼ ;58°, c ¼ ;�116°).

The hydrogen-bonding pattern surrounding the PMPcofactor in ArnB is displayed in Figure 4B. Like thatobserved in ColD, the phosphate group is anchored to theprotein through side chain hydroxyl groups and amidenitrogens and, in addition, is surrounded by severalordered water molecules. The main difference in thebinding of PMP to ArnB occurs at His 163, which, inColD is a glutamate. As observed in ColD, the indole sidechain of Trp 88 in ArnB forms a parallel stackinginteraction with the pyridoxamine ring of the cofactor.

Once the model for ColD was built and refined, a searchwith DALI (Holm and Sander 1996) revealed the closeststructural relative of ColD to be 3-amino-5-hydroxybenzoicacid synthase (ABHA synthase) rather than ArnB (Eadset al. 1999). ABHA synthase and ColD superimpose with anRMSD of 1.8 A for 677 structurally equivalent a-carbons.

Cook et al.




These two proteins display an even lower amino acidsequence identity of 18% amino acid, however. ABHAsynthase catalyzes the a,b-dehydration and 1,4-enolizationof 5-deoxy-5-amino-3-dehydroshikimic acid to produce3-amino-5-hydroxybenzoic acid, which serves as the starterunit for the production of ansamycin antibiotics (Ghisalbaand Nuesch 1981). A superposition of the polypeptidechains for ColD and ABHA synthase is presented in Figure5A. As noted for the comparison of the ColD and ArnBmodels, there are three loops (Phe 68–Arg 74, Lys 224–Val

230, and Lys 309–Ile 315) where ColD and ABHAsynthase differ to any significant extent. The cis peptidein ArnB is not conserved in ABHA synthase. A close-upview of the binding region for the internal aldimine inABHA synthase is depicted in Figure 5B. The hydrogenbonding pattern around the phosphate is a similar to thatobserved in both ColD and ArnB. As seen in ArnB,a histidine side chain lies within 2.8 A of O39 of thepyridoxamine ring. This residue varies among PLP-dependent enzymes, and has also been observed as an

Figure 3. The structure of ColD with a glutamate ketimine intermediate trapped in the active site. Electron density corresponding to

the ketimine intermediate is presented in A. The map was calculated in the same manner as described for Figure 2B. Note that the

a-carboxylate group of the intermediate is not well ordered, suggesting free rotation. A close-up view of the active site is displayed in

B. Possible hydrogen bonding interactions are indicated by the dashed lines. For the sake of clarity, the hydrogen bonding interaction

between Ser 183 and the phosphoryl group of the ketimine intermediate was omitted. Ser 183 forms the same interaction with the

ketimine phosphoryl group as it does with hydrated PLP (Fig. 2C).

Structure of ColD




Fig. 3 live 4/C

asparagine or aspartate. Trp 88 in ColD, which stacks withthe pyridoxal ring of the hydrated PLP cofactor, is replacedwith a phenylalanine in ABHA synthase.

A glutamate ketimine intermediate was first observedstructurally in chicken mitochondrial aspartate amino-transferase (AATase) (Malashkevich et al. 1993). Thepolypeptide chains for the dimers of ColD and AATasesuperimpose less well with an RMSD of 3.5 A for 531structurally equivalent a-carbons. Not surprisingly, thereare significant differences in the active sites of these twoenzymes. As can be seen in Figure 6, the conformationsof the trapped ketimine intermediates in ColD andAATase are quite different, with variations initiating atC49 of the cofactor. In AATase, the pyridoxamine ring ofthe cofactor and N, Ca, Cb, and the a-carboxylate groupof glutamate all lie nearly in a plane. Furthermore, theglutamate (or Schiff base) nitrogen sits within 2.5 A from

the pyridoxamine O39 in the AATase ketimine. This isindicative of a proton being shared between these twoatoms. Strikingly, the distance between O39 and the Schiffbase nitrogen in the ColD model is 3.5 A. Moreover, inColD, O39 lies at 2.7 A from the carboxylate of Glu 162,which means that there is a proton located somewherebetween them. In light of these differences, it can bespeculated that in the ColD ketimine intermediate a protonis not shared between O39 and the Schiff base nitrogen asis so often drawn in textbooks, but rather the proton isformally bonded to the nitrogen. This is referred to as theketoenamine tautomer, and it has been postulated thatenzymes can enhance their catalytic activity by pro-moting this tautomeric form (Toney 2001). The twoketimine intermediates observed in ColD and AATasediffer by ;77° with respect to the dihedral angle definedby C3–C4–C49–N. Other differences in the ketimine

Figure 4. Comparison of ColD with ArnB. A superposition of the a-carbon traces for ColD (purple) and ArnB (white) is presented in

A. Labeled residues are for ColD. The observed hydrogen-bonding pattern around PMP in ArnB is shown in B. X-ray coordinates for

this figure were obtained from the Protein Data Bank (accession no. 1MDO).

Cook et al.




Fig. 4 live 4/C

conformations seen in ColD and AATase include thepositioning of the a-carboxylate moiety. In AATase, itparticipates in a salt bridge with Arg 386, whereas inColD, it lies close to His 188.

It is important to note that the orientation of theglutamate ketimine intermediate observed in AATase (Fig.6) can be easily accommodated in the ColD active site if theside chain of Trp 88 moves slightly. In this orientation thea-carboxylate of the ketimine intermediate in ColD wouldparticipate in a salt bridge with the guanidinium group ofArg 331 and the side chain carboxylate would be situatednear the imidazole group of His 215. Importantly, this

orientation of the ketimine in the ColD active site places theside chain of His 188 on the re face of its sp2 hybridizeda-carbon. As will be discussed below, it has been postulatedthat bond cleavage at the sugar C39 and bond formation atthe cofactor C49 most likely occurs on the re face of theSchiff base complex (Alam et al. 2004). The unusualbinding of the glutamate ketamine observed in the ColDactive site, where the planarity of the intermediate isbroken, may simply be a function of the pH (6.0) at whichthe crystals were obtained.

While GDP was included in the initial crystallizationtrials, it was not observed in any of the electron density

Figure 5. Comparison of ColD with ABHA synthase. Shown in A is a superposition of the a-carbon traces for ColD (purple) and

ArnB (white). Labeled residues are for ColD. The observed hydrogen-bonding pattern around the internal aldimine in ABHA synthase

is shown in B. X-ray coordinates for this figure were obtained from the Protein Data Bank (accession no. 1B9H). Those residues

highlighted in green belong to one subunit, while Asn 234 is contributed by the second subunit in the dimer of ABHA synthase.

Structure of ColD




Fig. 5 live 4/C

maps. Repeated attempts at soaking crystals in solutionscontaining GDP up to 10 mM were also unsuccessful. Fromthe structure of the ColD dimer it is clear that a rather largeopen cleft extends from the ketimine intermediate to thesurface. Preliminary attempts at collecting X-ray data fromcrystals grown in the presence of GDP-4-keto-6-deoxy-mannose have, however, met with some limited success(unpublished data). In a map calculated with coefficients ofthe form (Fo � Fc), a tube of electron density was observedthat extends from the PLP cofactor toward the proteinsurface. Embedded in this density were two large peaks(4s) separated by ;2.8 A. Assuming that these two peakscorrespond to the phosphoryl moieties of the substrate, itwas possible to build a model for it with its sugar C49 ketogroup placed near the PLP cofactor and the phosphoryl andGDP-ribosyl groups extending toward the solvent. Impor-tantly, this model places the phosphoryl moieties near thepositively charged side chains of Arg 331 in subunit 1 andArg 219 and Arg 250 in subunit 2. If this model is correct,most of the ligand/protein binding interactions would beprimarily through the sugar and the phosphoryl groups ofthe substrate. It is not known whether ColD exhibits a strictspecificity for GDP or can utilize other nucleotide-linkedsugars.

ColD functions in the GDP–colitose biosynthetic path-way by removing the hydroxyl group at C39. Likecolitose, tyvelose is another 3,6-dideoxyhexose thatoccurs in the O-antigens of some Gram-negative bacteria.However, the removal of its 39-hydroxyl group occurs viaa completely different reaction mechanism. Specifically,the dehydration at C39 is catalyzed by the E1/E3 complex,which requires NAD, FAD, and [2Fe-2S] clusters (He and

Liu 2002). Indeed, ColD is an especially intriguing sugardehydratase in that it lacks any requirement for NAD,FAD, and iron–sulfur centers.

A proposed mechanism for ColD has been previouslyput forth (Alam et al. 2004), and is presented in adaptedforms in Schemes 2 and 3. In the first step (Scheme 2),PLP is converted to PMP through its reaction withL-glutamate. A catalytic base is required for this processwhich, based on the structure presented here, we proposeto be His 188. This residue resides within ;4.4 A of theglutamate ketimine a-carbon. If glutamate is modeledinto the active site on the basis of the observed ketimineorientation in AATase, His 188 is positioned such that itsimidazole ring is ;3 A from the a-carbon proton.

Following the production of PMP, a Schiff base is formedbetween the C49 sugar carbon and the cofactor, labeled 1 inScheme 3. From detailed biochemical investigations, it isknown that deprotonation at C49 of the cofactor is stereo-specific, and that a catalytic base removes the pro-Rhydrogen (Alam et al. 2004). Again, based on the ColDmodel presented here, it can be speculated that His 188serves as the catalytic base to initiate the deprotonation eventleading to the intermediate labeled 2 in Scheme 3. The ColDactive site is markedly devoid of potential bases other thanHis 188, His 215, Asp 159, and Glu 162. As previouslynoted, both Asp 159 and Glu 162 are involved in hydrogenbonding interactions with the cofactor ring, and most likelyare not involved in catalysis. Likewise, His 215 is not in thecorrect position to promote catalysis in light of previouslyreported stereochemical considerations (Alam et al. 2004).

The expulsion of the sugar C39 hydroxyl group, asindicated in 2, requires an active site acid. Most likely His

Figure 6. Comparison of the ColD glutamate ketimine intermediate with that observed in aspartate aminotransferase. Those residues

highlighted in blue bonds correspond to the ColD model, whereas those amino acids depicted in green bonds correspond to aspartate

aminotransferase (PDB accession no. 1MAQ). Residue labels beginning with an asterisk refer to aspartate aminotransferase.

Cook et al.




Fig. 6 live 4/C

188, which is now protonated, functions in this capacity.Following the departure of the C39 hydroxyl moiety in theform of water, a D3,4-aminomannoseen intermediate isformed as shown in 3. Hydrolysis of this intermediateleads to the loss of PLP and the intermediate shown in 4of Scheme 3. We propose that His 188 serves to activatea water molecule for this hydrolysis event. It has beendemonstrated that in the next step labeled 4 in Scheme 3,the incorporation of a hydrogen at the sugar C39 isstereospecific (Alam et al. 2004). Again, His 188 prob-ably functions in this step as the active site acid. It ispresently not known whether the final expulsion ofammonia, as indicted in 5 in Scheme 3, occurs on theenzyme or in solution.

Why did ColD evolve with a histidine in place of thelysine that normally forms an internal aldimine with PLP?We hypothesize that this occurred because of the uniquechemistry involved in the reaction. In going from 3 to 4 inScheme 3, an enzymatic base is required. In theory, thiscould be accomplished by a lysine residue. However, inthis case, the lysine would most likely react directly withthe C49 of the cofactor to form an internal aldimine. Assuch, the next step leading to 4, which is the stereospe-cific protonation at C39, would never occur given thearrangement of side chains in the active site of ColD. Byhaving a histidine residue in place of a lysine, theformation of an internal aldimine is prevented, andimportantly, histidine provides the functional group

Scheme 3.

Structure of ColD




Scheme 3 live 4/C

necessary for the reaction to proceed from 3 to 4.Experiments designed to test this hypothesis are presentlyunderway.

From a chemical viewpoint, the in vitro synthesis ofnucleotide-linked sugars is a complicated and arduousprocess at best. But nature has developed a large reper-toire of enzymes for the specific and efficient biosynthe-sis of unique carbohydrates. To better understand thedeterminants of protein-substrate specificities and thefundamental reaction mechanisms utilized in these bio-synthetic pathways, the structures and functions of theenzymes must be carefully studied. The models of ColDpresented here provide new and powerful structuralfoundations upon which to more fully explore thisfascinating PLP-dependent enzyme.

Materials and methods

Molecular cloning of the ColD gene

Genomic DNA from enterohemorrhagic E. coli O55:H7 wasprovided by the laboratory of Dr. Rodney Welch, Department ofMedical Microbiology and Immunology, University of Wiscon-sin–Madison. The colD gene was PCR-amplified using primersthat introduced a 59 NdeI site and a 39 XhoI site. The purifiedPCR product was A-tailed and ligated into the pGEM-T(Promega) vector for screening and sequencing. A ColD-pGEM-T vector construct of the correct sequence was thenappropriately digested and ligated into a pET28b(+) vector thathad been modified to provide an N-terminal TEV proteaserecognition site (Thoden et al. 2005).

DH5-a E. coli cells were transformed with the resultingplasmid and plated onto LB Agar Kan plates. Multiple colonieswere picked to grow overnight in LB Kan medium from whichplasmid DNA was isolated. Plasmids were tested for thepresence of the colD gene by digestion with NdeI and XhoI.

Protein expression and purification

The ColD-pET28JT plasmid was used to transform Rosetta(DE3) E. coli cells (Novagen). The culture was grown at 37°Cwith shaking until an optical density of 0.9 was reached at600 nm. The flasks were then cooled on ice and induced with1 mM IPTG. Cells were allowed to express at 16°C for 24 h afterinduction with IPTG.

The cells were harvested and disrupted by sonication on ice.The lysate was cleared by centrifugation and ColD was purifiedutilizing Ni-NTA resin (Qiagen) according to the manufacturer’sinstructions. The N-terminal His-tag was cleaved from theprotein by the addition of both 1 mM DTT and TEV proteasein a 1:50 (TEV protease:ColD) molar ratio. The cleavagereaction was carried out at 4°C for 36 h and subsequentlypassed through the nickel column to remove the TEV protease.The ColD protein sample was dialyzed against 25 mM Tris-HCland 200 mM NaCl (pH 8.0) with the addition of 1 mM PLP and2 mM glutamate or a-ketoglutarate. Following dialysis, thesample was concentrated to 20 mg/mL based on an extinctioncoefficient of 0.95 (mg/mL)�1 �cm�1 as calculated with theProtParam tool on the ExPASy Proteomics Server.

Crystallization of ColD

Crystallization conditions were first surveyed by the hangingdrop method of vapor diffusion with a sparse matrix screendeveloped in the laboratory. Large crystals were subsequentlygrown via batch methods by mixing 20 mL of a protein solutionwith 20 mL of precipitant. Protein concentrations were typicallyat 20 mg/mL and the solutions contained 25 mM Tris (pH 8.0),200 mM NaCl, 1 mM PLP, and 2 mM glutamate or a-ketoglu-tarate. The precipitant solutions contained 28% poly(ethyleneglycol) 3400, 100 mM MES (pH 6.0), 400 mM MgCl2, 1 mMPLP, 2 mM glutamate or a-ketoglutarate, and 1 mM GDP. Thecrystals grew to maximum dimensions of 1.8 3 0.5 3 0.5 mm in;2 d. They belonged to the space group P212121, with unit celldimensions of a ¼ 74.9, b ¼ 88.1, and c ¼ 125.8, and containeda dimer in the asymmetric unit.

Structural analysis of ColD

X-ray data sets were collected from two different crystalcomplexes. In one form the protein contained hydrated PLP inits active site, while in the second form the protein contained thetrapped glutamate ketimine intermediate. All X-ray data weremeasured with a Bruker HISTAR area detector system. TheX-ray source was CuKa radiation from a Rigaku RU200 X-raygenerator operated at 50 kV and 90 mA. The X-ray data wereprocessed with XDS (Kabsch 1993) and internally scaled withXSCALIBRE (I. Rayment and G. Wesenberg, unpubl.). Rele-vant X-ray data collection statistics are presented in Table 1.

The structure of the ColD/hydrated PLP complex was solvedvia molecular replacement with the program Phaser andemploying ArnB, a 4-amino-4-deoxy-L-arabinose lipopolysac-charide-modifying enzyme (PDB accession no. 1MDO) asa search model (Noland et al. 2002). The electron densitiescorresponding to the two subunits in the asymmetric unit wereaveraged with the software package DM (Cowtan and Main1998), and a model was manually built on the basis of thisaveraged map. The model was then placed back into the unit cellfor subsequent least-squares refinement with the softwarepackage TNT (Tronrud et al. 1987). Alternate cycles of manualmodel rebuilding and least-squares refinement reduced theR-factor to 17.0% at 1.8 A resolution. The structure of the trap-ped glutamate ketimine intermediate was solved via differenceFourier methods. Alternate cycles of manual model rebuildingand least-squares refinement of this model reduced the R-factor

Table 1. X-ray data collection statistics

Space groupP212121

(hydrated aldehyde)P212121

(ketimine intermediate)

Unit cell dimensions (A) a ¼ 74.9, b ¼ 88.1,

c ¼ 125.8

a ¼ 74.9, b ¼ 88.0,

c ¼ 126.0

Resolution limits (A) 30–1.8 (1.88–1.8)a 30–1.9 (1.99–1.9)

Number of independent

reflections 76,612 (8943) 72,854 (7334)

Completeness (%) 92 (85) 88 (79)

Redundancy 3.7 (1.5) 3.1 (1.4)

Avg I/Avg s(I) 9.9 (1.4) 11.3 (1.3)

Rsymb (%) 5.86 (32.3) 5.46 (31.0)

a Statistics for the highest resolution bin.bRsym ¼ (S|I � �I|/SI) 3 100.

Cook et al.




to 17.6% at 1.9 A resolution. Relevant refinement statistics aregiven in Table 2.

In each subunit (for both complexes) Met 193, Val 334, andAsn 354 adopt f, c angles that are outside of the allowedregions of the Ramachandran plot (all near the ‘‘nucleophileelbow’’ region of f ¼ ;60°, c ¼ ;�100°). The electrondensities for these residues are unambiguous. Met 193 ispositioned ;10 A from the phosphate of PLP, and is locatedin a sharp, nonclassical reverse turn. The side chain of Val 334lies within ;5 A of the PLP, and is situated in a Type II9 turn.Asn 354 sits in a surface loop ;20 A from the active site. Otherthan these amino acids, 89.3%, 10.5%, and 0.1% of theremaining residues fall into the ‘‘most-favored,’’ ‘‘additionallyallowed,’’ and ‘‘generously allowed’’ regions of the Ramachan-dran plot, respectively. Both complexes contain one magnesiumion. This cation resides near a crystalline lattice interface. While1 mM GDP was included in the crystallizations, it was notobserved in the electron density maps.

PDB accession numbers

X-ray coordinates have been deposited in the Protein Data Bankwith accession numbers 2GMS and 2GMU.

Acknowledgments

We thank Professors Ivan Rayment and W.W. Cleland forhelpful discussions. We are indebted to Professor Rodney Welchfor graciously providing genomic DNA. This research wassupported by a grant from the NIH (DK47814 to H.M.H.).

References

Alam, J., Beyer, N., and Liu, H.W. 2004. Biosynthesis of colitose: Expression,purification, and mechanistic characterization of GDP-4-keto-6-deoxy-D-

mannose-3-dehydrase (ColD) and GDP-L-colitose synthase (ColC). Bio-chemistry 43: 16450–16460.

Bastin, D.A. and Reeves, P.R. 1995. Sequence and analysis of the O antigengene (rfb) cluster of Escherichia coli O111. Gene 164: 17–23.

Beyer, N., Alam, J., Hallis, T.M., Guo, Z., and Liu, H.W. 2003. Thebiosynthesis of GDP-L-colitose: C-3 deoxygenation is catalyzed by a uniquecoenzyme B6-dependent enzyme. J. Am. Chem. Soc. 125: 5584–5585.

Cowtan, K. and Main, P. 1998. Miscellaneous algorithms for density modifi-cation. Acta Crystallogr Sect D 54: 487–493.

Eads, J.C., Beeby, M., Scapin, G., Yu, T.W., and Floss, H.G. 1999. Crystalstructure of 3-amino-5-hydroxybenzoic acid (AHBA) synthase. Biochem-istry 38: 9840–9849.

Edstrom, R.D. and Heath, E.C. 1965. Isolation of colitose-containing oligo-saccharides from the cell wall lipopolysaccharide of Escherichia coli.Biochem. Biophys. Res. Commun. 21: 638–643.

Ghisalba, O. and Nuesch, J. 1981. A genetic approach to the biosynthesis of therifamycin-chromophore in. Nocardia mediterranei. IV. Identification of3-amino-5-hydroxybenzoic acid as a direct precursor of the seven-carbonamino starter-unit. J. Antibiot. (Tokyo). 34: 64–71.

Harris, C.M., Johnson, R.J., and Metzler, D.E. 1976. Band-shape analysisand resolution of electronic spectra of pyridoxal phosphate and other3-hydroxypyridine-4-aldehydes. Biochim. Biophys. Acta 421: 181–194.

He, X.M. and Liu, H.W. 2002. Formation of unusual sugars: Mechanisticstudies and biosynthetic applications. Annu. Rev. Biochem. 71: 701–754.

Hisatsune, K., Kondo, S., Isshiki, Y., Iguchi, T., Kawamata, Y., and Shimada, T.1993. O-antigenic lipopolysaccharide of Vibrio cholerae O139 Bengal,a new epidemic strain for recent cholera in the Indian subcontinent.Biochem. Biophys. Res. Commun. 196: 1309–1315.

Holm, L. and Sander, C. 1996. Mapping the protein universe. Science 273:595–603.

Jansonius, J.N. 1998. Structure, evolution and action of vitamin B6-dependentenzymes. Curr. Opin. Struct. Biol. 8: 759–769.

Kabsch, W. 1993. Automatic processing of rotation diffraction data fromcrystals of initially unknown symmetry and cell constants. J. Appl.Crystallogr. 26: 795–800.

Kennedy, J.F. and White, C.A. 1983. Bioactive carbohydrates in chemistry,biochemistry, and biology.. Ellis Horwood Publishers, Chichester, UK.

Kotandrova, N.A., Gorshkova, R.P., Zubkov, V.A., and Ovodov Iu, S. 1989.The structure of the O-specific polysaccharide chain of the lipopolysac-charide of Yersinia pseudotuberculosis serovar VII. Bioorg. Khim. 15:104–110.

Lerouge, I. and Vanderleyden, J. 2002. O-antigen structural variation: Mech-anisms and possible roles in animal/plant-microbe interactions. FEMSMicrobiol. Rev. 26: 17–47.

Table 2. Least-squares refinement statistics

30–1.8 (hydrated aldehyde) 30–1.9 (ketimine intermediate)

Resolution limits (A)aR-factor (overall) %/number of reflections 17.0/71,402 17.6/61,081

R-factor (working) %/number of reflections 16.2/56,614 17.4/55,002

R-factor (free) %/number of reflections 21.3/6249 23.1/6079

Number of protein atoms 6265 6259

Number of hetero-atoms 389b 346c

Average B values (A2)

Protein atoms 33.2 38.7

PLP derivatives 27.6 36.9

Solvents 43.1 46.7

Weighted RMSDs from ideality

Bond lengths (A) 0.008 0.009

Bond angles (°) 1.9 2.6

Trigonal planes (A) 0.005 0.005

General planes (A) 0.017 0.016dTorsional angles (°) 17.2 17.5

aR-factor¼ (S|Fo�Fc|/S|Fo|)3100, where Fo is the observed structure-factor amplitude and Fc is the calculatedstructure-factor amplitude.b These include two hydrated PLP molecules, one magnesium ion, and 353 water molecules.c These include two ketimine intermediates, one magnesium ion, and 295 water molecules.d The torsional angles were not restrained during the refinement.

Structure of ColD




Luderitz, O., Staub, A.M., Stirm, S., and Westphal, O. 1958. Colitose,3-desoxy-L-fucose, a new sugar building block with immunospecificfunctions in the endotoxins, lipopolysaccharides, of some Gram-negativebacteria. Biochem. Z. 330: 193–197.

Malashkevich, V.N., Toney, M.D., and Jansonius, J.N. 1993. Crystal structuresof true enzymatic reaction intermediates: Aspartate and glutamate keti-mines in aspartate aminotransferase. Biochemistry 32: 13451–13462.

Muldoon, J., Perepelov, A.V., Shashkov, A.S., Gorshkova, R.P.,Nazarenko, E.L., Zubkov, V.A., Ivanova, E.P., Knirel, Y.A., andSavage, A.V. 2001. Structure of a colitose-containing O-specific poly-saccharide of the marine bacterium Pseudoalteromonas tetraodonis IAM14160(T). Carbohydr. Res. 333: 41–46.

Nishino, J., Hayashi, H., Ishii, S., and Kagamiyama, H. 1997. An anomalousside reaction of the Lys303 mutant aromatic L-amino acid decarbox-ylase unravels the role of the residue in catalysis. J. Biochem. 121: 604–611.

Noland, B.W., Newman, J.M., Hendle, J., Badger, J., Christopher, J.A.,Tresser, J., Buchanan, M.D., Wright, T.A., Rutter, M.E., Sanderson, W.E.,et al. 2002. Structural studies of Salmonella typhimurium ArnB (PmrH)aminotransferase: A 4-amino-4-deoxy-L-arabinose lipopolysaccharide-modifying enzyme. Structure 10: 1569–1580.

Rupprath, C., Schumacher, T., and Elling, L. 2005. Nucleotide deoxysugars:Essential tools for the glycosylation engineering of novel bioactivecompounds. Curr. Med. Chem. 12: 1637–1675.

Silipo, A., Molinaro, A., Nazarenko, E.L., Gorshkova, R.P., Ivanova, E.P.,Lanzetta, R., and Parrilli, M. 2005. The O-chain structure from the LPS ofmarine halophilic bacterium Pseudoalteromonas carrageenovora-typestrain IAM 12662T. Carbohydr. Res. 340: 2693–2697.

Thoden, J.B., Timson, D.J., Reece, R.J., and Holden, H.M. 2005. Molecularstructure of human galactokinase: Implications for Type II galactosemia.J. Biol. Chem. 280: 9662–9670.

Toney, M.D. 2001. Computational studies on nonenzymatic and enzymaticpyridoxal phosphate catalyzed decarboxylations of 2-aminoisobutyrate.Biochemistry 40: 1378–1384.

———. 2005. Reaction specificity in pyridoxal phosphate enzymes. Arch.Biochem. Biophys. 433: 279–287.

Tramonti, A., John, R.A., Bossa, F., and De Biase, D. 2002. Contribution ofLys276 to the conformational flexibility of the active site of glutamatedecarboxylase from Escherichia coli. Eur. J. Biochem. 269: 4913–4920.

Tronrud, D.E., Ten Eyck, L.F., and Matthews, B.W. 1987. An efficient general-purpose least-squares refinement program for macromolecular structures.Acta Crystallogr A 43: 489–501.

Weymouth-Wilson, A.C. 1997. The role of carbohydrates in biologically activenatural products. Nat. Prod. Rep. 14: 99–110.

Xiang, S.H., Haase, A.M., and Reeves, P.R. 1993. Variation of the rfb geneclusters in Salmonella enterica. J. Bacteriol. 175: 4877–4884.

Zhang, X.-J. and Matthews, B.W. 1995. EDPDB: A multifunctional tool forprotein structure analysis. J. Appl. Crystallogr. 28: 624–630.

Cook et al.




Documents

The structure of GDP-4-keto-6-deoxy-d-mannose-3 ... · pyridoxamine 59-phosphate (PMP) by reacting with gluta-mate as indicated in Scheme 2. In the first step, glutamate forms a Schiff